-
TRANSPORTATION RESEARCH RECORD 1288 59
Stabilization Characteristics of Class F Fly Ash
MuMTAZ A. UsMEN AND JOHN J. BowDERS, JR.
Stabilized fly ash is a rnix:ture of fly a h and lim . or fl y a
·h and cement. c mpact d a l optimum moisture contt.:nt and cured
10 form a product-like . oil-lime or ·oil-cement. Limited past
appli-cation and engineering propertie or tabilized class F fly a h
are di cussed. A research study wa undertaken t establi h the phy ·
ical, chemical, compaction, strength, and durability
characteris-tic o . class F fly ash rnbilized with lime, cement, or
lime/cement combinations. Two ashes obtained from West Virginia
power plants were included in the laboratory te ting pr gram . ll
wa found that although the ashes arc quite diffc rclll in prop rtie
, b th a he · can be succes fully stal ilized to produce pozzolanic
mixtures of adequate strength and durability for use as ba or
liner, with the addition of a proper amount of stabi lizer and by
allowing the mixture to cure for a suffici.emly Jong period. ement
stabilization. in general , produced helter trength and durability
than lime stabi liza tion for a given ·tabilizer content f r curing
periods up to 56 days. Fre ze-thaw cycles cau ed sub tantial
strength losses, and wet-dry cycle · resulted in trength ga ins.
Vacuum saturation with water and an acetic acid solution produced
inter-mediate effects. Very good correlations were found between
freeze-thaw and water vacuum saturation tests.
Large quantities of fly ash continue to be generated by
coal-burning power plants, and the disposal of this material in a
safe, economical, and environmentally acceptable manner is becoming
increasingly troublesome for the electric utility industry and
becoming a public concern. The most desirable way of disposal is
utilization, which provides economic ben-efits by reducing disposal
costs and mitigates possible negative environmental effects through
proper engineering controls. Fly ash has been used in many types of
engineering appli-cations because of its wide availability and
desirable pozzo-lanic (and self-hardening) characteristics and has
been used as an admixture in cement and concrete and as a
stabilizing agent (in combination with lime or cement) for soils
and aggregates in pavement subgrades, bases, and subbases. Fly ash
also has been used as a fill material on a limited scale.
The use of fly ash in concrete and aggregate/soil stabili-zation
applications has proven beneficial both technically and
economically , but relatively small amounts of fly ash can be
exploited in those types of projects because, in most cases, fly
ash constitutes a small percentage of the total material
composition . This suggests that, in view of the economic and
environmental concerns mentioned, further benefits and incentives
remain for establishing utilization schemes that will incorporate
larger amounts of ash . One such scheme is "sta-
M. A. Usmen , Department of Civil Engineering, Wayne State
University, Detroit , Mich. 48202. J . J. Sowders , Jr., Department
of Civil Engineering, West Virginia University, Morgantown, W. Va.
26506-6101.
bilized fly ash," defined here as a pozzolanic mixture of fly
ash and lime or cement compacted at optimum moisture con-tent to
form a product-like soil-lime or soil-cement that can serve as a
base or subbase course for pavements or as a low permeability liner
or cut-off material when designed (pro-portioned) to meet pertinent
performance criteria. Because this material does not contain any
aggregate or soil , the use of fly ash is maximized per ton or
cubic yard of base , sub base , or liner material constructed. Fly
ash in such an application serves the dual role of pozzolan and
aggregate.
Detailed technical information is not available on stabilized
fly ash although an abundance of information exists on the
technology for pozzolanic base courses employing mixtures of lime,
fly ash, and aggregate (LFA); cement, fly ash, and aggregate (CFA);
and lime, cement, fly ash, and aggregate (LCFA) (1-3), as well as
the use of fly ash in soil stabilization (3 ,4). A research study
was performed to review and docu-ment the limited existing
information from the literature on material properties and
applications and to produce new infor-mation on material properties
through an organized labora-tory study. Two class F (bituminous
coal based) fly ashes from West Virginia were included in the
laboratory study. Those ashes were first characterized by
subjecting them to standard ASTM tests for pozzolans . Next, the
ashes were mixed with hydrated lime and portland cement at varying
stabilizer con-tents to investigate compaction characte ristics
(optimum moisture content and maximum dry density) . Then, the
spec-imens were fabricated, cured for different lengths of time ,
and tested for strength and durability. Findings of those
inves-tigations are reported in this paper. Permeability and
leachate characteristics of the stabilized fly ash mixtures were
also studied as part of this research project. However, relevant
information and findings concerning those aspects are pre-sented
elsewhere (5) and are discussed by Bowders et al. in a companion
paper in this Record .
PAST RESEARCH AND UTILIZATION
Material Properties
It is known that the most unique and outstanding
charac-teristics of fly ash are pozzolanic reactivity and being
self-hardening. Pozzolanic reactivity relates to the ability of fly
ash to form cementitious products at ordinary temperatures when
combined with alkali and alkaline earth hydroxides in the presence
of moisture. The alkali and alkaline earth hydroxides needed to
achieve pozzolanic reactions are pro-vided by adding lime or cement
to fly ash. If they are internally
-
60
present in sufficient amounts (e.g., as CaO or MgO), then the
fly ash exhibits self-hardening behavior in addition to
pozzolanicity. Fly ash is designated as class F or class C
depending on the parent coal source. Class Fash is derived from
bituminous or anthracite coals burned mostly in the eastern,
midwestern, and southern United States. Class Cash comes from
subbituminous or lignite coals predominantly mined in the western
United States. It is almost always necessary to combine class F tly
ash with lime or cement to produce poz-zolanic reactions, but this
may not be needed with class C ashes, which contain significant
amounts of CaO. However, many class C ashes also produce better
stabilization charac-teristics when lime 01 cement is auueu
(3,6).
Pozzolanic reactions between lime and fly ash are complex.
According to Minnick (7), those reactions involve various
combinations of the hydrated calcium or magnesium in lime with the
amorphous silica and alumina in fly ash or both . Reaction products
may include tobermorite (calcium silicate hydrate), ettringite
(high-sulfate calcium sulfoaluminate: 3Ca0 Al20 3 CaS04 12H20), and
the low sulfate form of calcium sulfoaluminate (3Ca0 Al20 3 CaS04
12 H 20). The extent and rate of the reactions will be affected by
the fineness and chem-ical composition of fly ash, the type and
amount of stabilizer, moisture content, temperature, and age. When
cement is used in lieu of lime to stabilize fly ash, it hydrates
relatively quickly on contact with moisture and produces its own
cementitious compounds and also releases some free lime that can
further react with fly ash in a pozzolanic manner. Consequently ,
cement enhances short-term strength.
Limileu information has been reported in the literature on the
strength characteristics of lime- and cement-stabilized fly ash
(6,8-10). Evident from those studies for mixtures of lime and fly
ash is that normal (70°-75°F) cured unconfined com-pressive
strengths at 7 to 28 days range from 100 psi to 1200 psi and that
longer curing periods (90 days and over) may yield strengths
exceeding 2000 psi. Cement may produce two to three times higher
strengths in the short term, but the differences largely disappear
in the long term.
Durability data on stabilized fly ash are very limited. Gray and
Lin (9) have found that both lime and cement stabilization
drastically reduce the frost susceptibility of fly ash. Freeze-thaw
durability evaluations by Joshi et al. (JO) have revealed that
mixtures of lime and fly ash have questionable durability in the
short term and that mixtures of cement and fly ash produce
satisfactory results .
Complete mix-design data for stabilized fly ash could not be
found in the literature . Therefore, effects of stabilizer
con-tents on mixture properties cannot be clearly assessed.
Evi-dence exists that increased cement content will increase
mix-ture strength (I 1). However, strength may increase with
increasing lime content (10,11) or results may be varied (9) .
According to GAI Consultants (6), the standard soil-cement
wet-dry and freeze-thaw durability tests, using the brushing
technique, are not suitable for cement-stabilized fly ash. Because
wet-dry cycles apparently produce negligible effects on dura-bility
and freeze-thaw cycles are unduly abrasive, and because test
results are dependent on sample preparation techniques, it is
suggested that compressive strength tests can be used alone for
ensuring adequate durability. For cement-stabilized fly ash, a
7-day normal cured strength between 400 and 800 psi has been
specified, along with the requirement that the
TRANSPORTATION RESEARCH RECORD 1288
strength of the mix must increase with time. A minimum 28-day
normal cured strength of 550 psi has been recommended for
lime-stabilized ash, again with the additional stipulation that
there must be strength increase with time. However, stabilized fly
ash pavements subjected to extreme service con-ditions should be
tested for durability by observing residual strength after a
suitable number of freeze-thaw cycles or after vacuum
saturation.
In general , a partial or full replacement of lime by cement in
pozzolanic mixtures has been considered advantageous, although this
has so far been tried only on pozzolan-aggregate mixes (2). In
addition to better early strength, cement appar-ently also enhances
durability . It has been suggested that the designer can have
better control over the mixture quality (by adjusting cement
content) and that nonspecification ashes not highly lime reactive
may be effectively stabilized with cement.
Applications
Cement-stabilized fly ash has been successfully used as a base
course material in England and France for many years and has been
specified and accepted on both public roads and private projects
(4,6). This type of application is relatively new in the United
States but is expected to gain increased attention especially in
locations where fly ash can economi-cally compete with alternative
aggregate materials. Field trials and demonstration projects have
been undertaken in recent years to evaluate the performance of
stabilized fly ash pave-ments (4,10-13). Three cases related to
cement-stabilized class F fly ash mixes are briefly described
here.
• In September 1975 a parking lot pavement consisting of an
8-in.-thick cement-stabilized fly ash base, which was topped by a
3-in. bituminous wearing surface, was constructed at Harrison Power
Station in Haywood, West Virginia (4) . The purpose of the project
was to demonstrate cement-stabilized fly ash as an easily
constructed and highly serviceable base cou1se. Cement and fly ash
were premixed with water in a pugmill at the rate of 83 lb of fly
ash and 10 lb of cement per cubic foot of compacted mix. An average
in-place density of 98.5 percent of the maximum Standard Proctor of
92.5 pcf was obtained at an optimum water content of 14 percent.
Average unconfined compressive strengths of cores taken from the
completed base course at 7 and 90 days were 566 and 869 psi,
respectively. Strengths of cores taken after a period of 180 days,
which encompassed a severe winter, indicated that the pavement had
experienced no strength loss. The parking lot has continued to
perform well.
• Cement-stabilized fly ash was used in the construction of a
base course for a haul road near American Electric Power's Clinch
River power plant in southwestern Virginia (12) . The
cement-stabilized fly-ash base course was designed by the
procedures presented by GAI Consultants ( 4,6). The resulting
pavement consisted of cement-treated fly ash base course 5.5 in.
thick overlain by a 1.5-in.-thick emulsified asphalt-stabilized
bottom ash surface course. A cement content of 14 percent of the
dry weight of the fly ash and a water content of 17 percent were
selected for the base-course mix. The haul road was subjected to a
low traffic volume, although many of the vehicles were heavy
trucks, and the road performed satisfactorily for several
years.
-
Usmen and Bowders
•An experimental project , using a 10-in.-thick cement
sta-bilized fly ash as the road base material, was conducted by the
Oakland County Road Commission in Michigan in 1983 (13). The mix
contained 10 percent cement and presented no unique problems during
pugmilling and construction. Ease of construction was roughly
comparable to the installation of a gravel or bituminous base.
Traffic control was also similar, except for a recommendation that
the traffic be kept off the completed base during the initial
curing period (7 days). Cores taken at 7 and 28 days yielded
unconfined compressive strength of 190 psi and 142 psi,
respectively. However, early laboratory testing on a sample of fly
ash from the plant indicated the design mix would produce a 7-day
strength over 400 psi. This difference probably arose from
discrepancies between labo-ratory testing conditions and in-place
field conditions. This level of performance, however, was
considered unsatisfac-tory, and this mixture was not recommended
for use in the Detroit, Michigan, area.
RESULTS OF THE LABORATORY STUDY
The laboratory studies involved testing of fly ashes in
unsta-bilized and stabilized form. The purpose of the study was to
determine the specification conformity of the ashes, the
com-paction behavior and parameters of the ash-stabilizer mixtures
in the freshly mixed condition, and the strength and durability
characteristics of the same mixes after curing. All testing in the
laboratory was performed on duplicate specimens to obtain average
results. If the test results had significant variability, then
additional tests were performed before averaging.
Materials
Two fly ashes were used in this study: Harrison, obtained from
the Harrison Power Plant in Haywood , West Virginia, and Amos,
acquired from the Amos Power Plant, in Nitro, West Virginia. The
samples were collected from the dry hop-pers in the power plants
and were transported to the labo-ratory for testing. The hydrated
lime used in this study was manufactured by the Greer Plant of
Morgantown, West Vir-ginia, and the Type I portland cement was
produced in Arm-strong, West Virginia. Both were bought in paper
sacks from local suppliers.
Ash Properties
A variety of ASTM specification tests were performed on the fly
ashes and included specific gravity (ASTM 0854); fine-ness, as
established by the amount retained when wet-sieved on No. 200 and
No. 325 sieves (ASTM D422); pozzolanic activity index with portland
cement and pozzolanic activity index with lime (ASTM C311); and
lime-pozzolan strength development (ASTM C593). A summary of the
test results is presented in Table 1, along with the related ASTM
specifi-cation criteria. Data on the chemical analyses of the ashes
shown in the table were provided by the utility companies, except
for the loss on ignition values and CaO contents, which were
determined in the laboratory.
61
The specific gravity values presented in Table 1 indicate that
the Harrison ash is much heavier than the Amos because of its high
Fe20 3 content. The Amos ash conversely has a higher total amount
of glassy components (Si02 , Al~03 , and Fe20 3 ) than does the
Harrison and has a higher pozzolanic activity index with lime and a
higher lime-pozzolan strength development value than does the
Harrison. However, both ashes exhibit excellent pozzolanic
reactivity with both cement and lime. The sieve analysis results
indicate that the Harrison is somewhat finer than the Amos. The
loss on ignition values are comparable for both ashes, with the
Harrison slightly lower. The values presented in the table indicate
that rela-tively small amounts of carbon and other combustible
mate-rials exist in the ashes. The Cao percentage for the Harrison
is significantly higher than that for the Amos. Overall, both
ashes, although quite different in properties, satisfy the ASTM
specification criteria for class F fly ashes for use in cement and
concrete and for lime-pozzolan stabilization.
Compaction Characteristics
Compaction characteristics of mixtures of fly ash and lime and
of fly ash and cement were investigated by performing Standard
Proctor tests (ASTM 0698) on materials by using varying stabilizer
contents . The maximum dry density (MOD) and optimum moisture
content (OMC) were obtained on each mixture . Results are presented
graphically in Figures 1 and 2 for Harrison lime (HL), Harrison
cement (HC), Amos lime (AL), and Amos cement (AC) mixtures. Results
for unsta-bilized mixtures (zero percent stabilizer) are also
included. Some differences exist between the compaction
characteristics of the two fly ashes. The Harrison with the higher
specific gravity produces higher maximum dry densities when
com-pared with the Amos. However, the Amos, being a lighter weight
material, yields higher optimum moisture contents because of the
larger surface area it has per unit mass.
The data for both ashes also indicate that increased lime
content results in increased OMC and decreased MOD, which can be
attributed to the fineness and light weight of lime. Conversely,
increased cement content does not appear to pro-duce any clear
trends, or any significant variation, relative to OMC and MOD. The
moisture-density relationships for individual mixtures were very
straightforward to obtain, and the standard laboratory procedures
posed no problems or anomalies.
Strength Development
The stabilized fly ash mixtures were first compacted in Proctor
molds at their OMC. They were then extracted from the molds and
placed in plastic closeable bags and cured in a moist room at 73°F
and 100 percent relative humidity. The mixtures were then tested
after specified curing periods for unconfined com-pressive strength
in the unsoaked condition to assess the degree of stabilization
through progressing pozzolanic reactions between the fly ashes and
the stabilizers. The soaking pro-cedure normally employed to
determine design strength was omitted to avoid the possibility of
negating effects that would obscure the results. (The soaking
procedure, however, was
-
62
TABLE 1 PROPERTIES OF FLY ASHES
Property
Specific Gravity
%Retained #200 Sieve
Fineness, % retained on #325 Sieve
Moisture Content (%)
Harrison Ash
2.81
4.4
14.4
0.1
Pozzolanic Activity Index 97.6 with cement (%)a
Pozzolanic Acti~ity Index 944 with lime (psi)
Lime-Pozzolan Strength 644 Development (psi)c
Silicon Dioxide (Si02) I % 34
Aluminum Oxide (Al 2 o3 ), % 21
Ferric Oxide (Fe2 o3 ), % 24
Sum of Sio2 , Al 2o3 , and
Fe2o3 , % 79
Loss on Ignition (%) 2.2
Cao (%) 6.8
Amos Ash
2 . 25
8.9
22.4
0.1
86.0
1003
979
58
30
4
92
2.5
1. 4
TRANSPOR TA TION RESEA RCH RECORD 1288
ASTM Specifications
ASTM C593 30.0 max.
ASTM C618 34.0 max.
ASTM C618 3.0 rnax.
ASTM C618 75 min.
ASTM C618 800 min.
ASTM C593 600 min.
ASTM C618
ASTM C618 12 rnax.
a - Cured l day at 73 F plus 27 days at 100 F
b - cured l day at 73 F plus 6 days at 130 F
c - Cured 7 days at 130 F
replaced by vacuum saturation, which is reported in the next
section.)
The different ash-stabilizer combinations and curing periods
employed in this phase of the study and the test results are
presented in Table 2. Both ashes were stabilized with 3, 6, 9, 12,
and 15 percent lime and with 3, 6, 9, 12, and 15 percent cement. In
addition, the Harrison was stabilized with 9 per-cent cement and 3
percent lime, 6 percent cement and 6 percent lime, and 3 percent
cement and 9 percent lime, to study the effects of using combined
stabilizers on mixture strength development . The Amos mixtures
were tested after 7 and 28 days of curing only, and the Harrison
mixtures were tested after 7, 28, and 56 days to assess the effects
of longer-
term curing. Unconfined compressive strengths for unstabil-ized
ashes (zero percent lime or cement) were also obtained to establish
baseline values.
From the results presented in Table 2, increasing cement content
causes considerable increases in the strength of both ashes for all
curing periods. Increasing lime content, however , may increase or
decrease strength. A slight decrease is observed with the Harrison,
in general with increasing lime contents at 7 and 28 days. However,
the trend reverses at 56 days. This may be caused by unfinished
pozzolanic reactions between lime and fly ash in the short term. In
the Amos lime mixtures , increased lime content causes negligible
strength gain at 7 days, but extended curing effects a notable
increase in strength .
-
Usmen and Bowders
(:; c. cl 100 c •HC ::.
o HL
1>.AC
A AL
Stabilizer Content, %
FIGURE 1 MDD versus stabilizer content fo r ash mixtures.
Overall, the results show that cement-stabilized ashes exhibit
higher strengths than their lime-stabilized counterparts,
regardless of the length of curing period. However, as the curing
period gets longer the differences between the cement-and
lime-stabilized fly ash mixes at a given stabilizer content become
somewhat smaller.
When comparing the strengths of the two ashes, the Har-rison has
developed higher strengths for both lime and cement stabilization.
This, unfortunately, cannot be readily predicted from the
pozzolanic reactivity test results presented in Table 1, because
the Amos appears to have higher pozzolanic reac-tivity with lime
than does the Harrison. However, the test results presented in
Table 1 are based on accelerated curing at high temperatures.
Further, the high CaO content and the fineness of the Harrison can
be important factors that con-tribute to high strength. The results
in Table 2 indicate that the Harrison ash shows very satisfactory
strength values with both lime and cement for all stabilizer
contents and curing periods. The Amos exhibits relatively low
strengths with lime in the short term, but extending curing results
in appreciable strength gains and reaches satisfactory levels at
higher lime contents, a favorable characteristic. The length of
curing, actually, has a very dramatic effect on all mixtures. The
longer the curing period, the higher the strengths.
Finally, the dominance of cement in the strength devel-opment of
stabilized fly ash mixes is quite evident from the results given in
Table 2 for the Harrison stabilized with com-bined lime and cement.
As the cement/lime ratio increases in
63
20
*-u" ::. 0 A AL
18 1>.AC OHL
•HC
16
0 6 12 Stabilizer Content , %
FIGURE 2 OMC versus stabilizer content for ash mixtures.
the combined stabilizer, the strength of the total mixture also
tends to increase. The mixtures containing combined stabi-lizers
indicate strengths much higher than those stabilized with lime
only. When compared with cement-stabilized mixtures, the
differences are much less and become quite insignificant at high
cement/lime ratios.
Durability Evaluation
Durability of lime- and cement-stabilized fly ash mixtures was
evaluated by obtaining the residual strength q, of the cured
specimens after subjecting them to different exposure con-ditions
and then comparing this value to the original (pre-exposure)
strength q0 • Three levels of stabilizer contents were used in the
preparation of the mixes: 3, 9, and 15 percent. However, part of
the Amos specimens was prepared only with 9 percent stabilizer to
economize on time and materials. Two curing periods , 7 and 28
days, were selected for specimen preparation.
Five types of exposure conditions were chosen for durability
evaluations. The first series of tests employed vacuum satu-ration
with water and was performed in accordance with ASTM C593. The
second series used a freeze-thaw cycles exposure and was performed
as outlined in ASTM D560, except for employing 10 cycles instead of
12 and for substituting uncon-fined compressive strength testing at
the end of the exposure period for the brushing and weighing
procedures. A similar
-
64 TRANSPORTATION RESEARCH RECORD 1288
TABLE 2 COMPRESSIVE STRENGTHS OF STABILIZED ASHES
Compressive Strength (psi) Mixture 7-day
Unstabilized Harrison 80
Harrison + 3% Lime 559 + 6% Lime 543 + 9% Lime 439 +12% Lime 394
+15% Lime 372
Harrison + 3% Cement 193 + 6% Cement 442 + 9% Cement 917 +12%
Cement 1074 +15% Cement
Harrison + 9% Lime and 3% Cement
Harrison + 6% Lime and 6% Cement
Harrison + 3% Lime and 9% C.ement
Unstabilized Amos
Amos + 3% Lime + 6% Lime + 9% Lime +12% Lime +15% Lime
Amos + 3% Cement + 6% Cement + 9% Cement +12% Cement +15%
Cement
approach was adopted in the third series, which involved wet-dry
cycles. In this case , the procedures specified in ASTM D559 were
modified so as to include 10 cycles of wetting and drying rather
than 12 and compressive strength testing instead of brushing and
weighing. The fourth and fifth series of dura-bility testing
involved a vacuum saturation exposure, this time using an acetic
acid solution (0.575 M with a pH of 2.5) instead of water. The
vacuum saturation period was extended to 2 hours for both series to
ensure better pervasion of the ;icid solution into the specimens.
This procedure was devised to test the durability of the mixtures
in an acidic environment (i.e., as a landfill liner).
Two sets of specimens were tested, one directly after vac-uum
saturation, another after 48 hours, to assess the effects of
prolonged exposure while the specimens were sealed in plastic bags.
The weights of all specimens were monitored throughout the
durability testing program to study moisture changes and material
loss owing to possible sample deterio-
1341
694
756
959
40
104 116 126 130 142
217 307 488 637 826
28-day 56-day
95
909 1116 780 1182 761 1220 895 1277 758 1138
422 756 792
1209 1353 1675 1773 1803 2423
1251 1695
1361 1687
1635 1928
40
148 239 360 482 669
251 508 764
1112 1492
ration. Visual observations were also performed to supple-ment
quantitative evaluations.
Durability test results for the mixtures and exposure
con-ditions are summarized in Tables 3 and 4 for 7 and 28 day
curing periods, respectively . Results are presented in terms of q,
values and q )q0 , which is the ratio of the residual strength and
the original strength. The original strength values used in
computing these ratios are those given in Table 2 for the same
mixtures. A q,lq" ratio of less than 1.0 indicates a strength
decrease as a result of the exposure, and a q, lq" ratio of greater
than 1.0 signifies a strength increase as a result of the
expo-sure. If any of those exposures and testing procedures are
adopted for durability evaluations, then the particular crite-rion
to be met by those parameters should be determined for the
anticipated field conditions to which the pavement or liner
material will be subjected. Low q, values and q,lq" ratios might
indicate a need to critically evaluate the potential dura-bility
problems for the given case.
-
TABLE 3 DURABILITY TEST RESULTS FOR STABILIZED ASHES (7
DAYS)
Exposure Cond1l1on Vac. Saturation Freeze-Thaw Wet-Dry Vac.
Saturat1on w1th Acetic Acid
wilil Waler Cycles Cycles Tested Tested after Imma1l lal a J :I!
!Ill ll!lU[S
Mixture a q /q l)
q/qo q/qo q/qo q/qo qr r " Q,_ qr qr Qr
(psi) (psi) (ps1) (ps1) (psi)
Herr1son + 3, L1me 446 O.BO '.17 3 0.67 1540 2.75 394 o. 70 400
0.72 + 9, L1me 346 0,79 129 0.27 2666 6.07 359 0.82 320 o. 73 +15,
L1me 317 0.85 56 o. 15 2845 7. 67 276 0. 74 283 o. 16
Harr1son + 3, Cement 1 fi9 0.88 295 1. 53 812 4.21 241 1. 25 103
'1.05 + 9, Cement 7:16 1.00 844 1. 14 3462 4.69 1166 2.09 1096 1.
96 +15, Cement 1069 1. 03 1106 1. 07 2348 2.26 977 1. 33 858
1.17
Amos + 3, L1me 64 0.62 0 0 c
+ 9, L1me 60 0.48 26 0.21 1313 10.42 64 0.51 58 0.46 +15, L1me
86 0.61 34 0.24
Amos + 3' Cement 155 o. 71 54 0.25 + 9' Cement 346 0.71 213 0.44
1393 2.85 201 o. 41 243 0.50 +15' Cement 585 o. 71 549 0.66
a - qr = Res1dual compressive strength (post-exposure) b - q" =
Or1g1nal compressive strength (pre-exposure) c - 1nd1cates tilal
test was not performed
TABLE 4 DURABILITY TEST RESULTS FOR STABILIZED ASHES (28
DAYS)
Vac. Saturation w1th Acet1c Acid Exposure Cond1t1on Vac.
Saturation Freeze-Thaw Wet-Ory Tested Tested after
w 1th Water C:z-cles Cycles rmmediatel}' 48 HOU(S
M1xture a
q/qo b
Q, q/qo qr q/qo Qr q/qo qr q/qo qr
(psi) (psi) (psi) (psi) (psi)
Harrison + 3, Lime 718 o. 79 690 o. 76 1791 1. 97 527 0.58 495
0.54 + 9, Lime 756 0.99 374 0.49 3084 4.05 541 o. 71 549 0 . 72
+15, L 1rne 611 0.81 362 0.48 3064 4. 04 543 ll. 72 573 0. 76
Harri son + 3' Cement 291 0.69 304 o. 72 1055 2 . 50 605 1. 43
645 1.53 + 9.' Cement 798 0.66 614 0.51 2646 l. 19 17 17 I. 42 1631
1. 35 +15, Cement 1202 0.67 868 0.48 38 70 l. 15 1733 0.96 2171 1.
20
c Amos + 3, Lime 90 0.61 28 0. 19
+ 9X Lime 229 0.64 36 0 . 10 1811 5 . 03 201 0.56 185 0. 51 +
15X Lime 374 0.56 44 0.07
Amos + 3X Cement 183 0.73 121 0.48 + 9X Cement 655 0.86 691 0.40
1353 I. 77 584 0. 76 603 0.79 +1511: Cement 1212 0.81 1431 0.96
a qr Residual compressive strength (post-exposure)
b - q 0
Orig1nal compressive strength (pre-exposure)
c 1nd1cates that test was not performed
-
66
Several observations can be made from the data presented in
Tables 3 and 4. First, stabilized mixtures of the Harrison ash have
produced much better durability in most cases than the stabilized
Amos mixtures. Residual strengths are invari-ably higher for the
Harrison primarily because the original strengths were higher to
start with and underscores the impor-tance of obtaining
sufficiently high strength in stabilized ash mixtures before their
exposure to possible detrimental service environments . The
Harrison has also produced higher q, lq11 ratios with few
exceptions. Cement-stabilized Harrison, in particular, shows
excellent durability with respect to all expo-sures, with an
exception observed in the q,)q0 ratios for the freeze-thaw test.
Cement-stabilized Amos may have done bet-ter than lime-stabilized
Amos, producing satisfactory dura-bilities in many cases,
particularly at relatively higher cement contents (greater than 9
percent) and longer curing periods (28 days). Lime-stabilized Amos
also has performed better after the 28-day curing period when
compared with the 7-day curing. This is true for most exposures .
However, this mixture failed in freeze-thaw after both curing
periods .
Overall, increased stabilizer contents and extended curing
periods enhance the durability of the stabilized ash mixtures, and
cement-stabilized mixtures perform better in most of the durability
exposure conditions. The freeze-thaw cycles test pro-duces the
severest exposure and results in substantial strength losses in
most cases. The wet-dry cycles test, however, invar-iably results
in very high strength gains for the specimens,
2000
1950
1900
1850
1800
"' E "' 0 E 1750 Cl
"iii 3:
1700
1650
1600
1550
1500 0 2 3 4 5 6 7 8 10
Cycle Number
FIGURE 3 Weights of lime-stabilized ashes exposed to freeze-thaw
cycles (28 days).
TRANSPOR TA TION RESEARCH RECORD 1288
indicating that this type of exposure will not be critical in
terms of the durability evaluation of stabilized fly ash.
Weight changes of 28-day cured specimens during the freeze-thaw
and wet-dry cycles are presented graphically in Figures 3, 4, 5,
and 6 to augment the results given in Table 4. The HL, HC, AL, and
AC symbols used in those figures denote Harrison lime, Harrison
cement, Amos lime, and Amos cement mixtures, respectively, and the
numbers at the end of each symbol designate the stabilizer
contents. Similar curves were obtained for 7-day curing. Evident
from Figures 3 and 4 is that both lime- and cement-stabilized ashes
have gained significant amounts of moisture after the first
freeze-thaw cycle. Afterward, an approximately constant weight is
main-tained for cement-stabilized ashes, but moisture gains
con-tinue in varying degrees in the lime-stabilized ashes. Weight
losses observed in the higher cycles indicate material loss owing
to cracking, sc
-
Usmen and Bowders
1860
1800
1740
1680
v
1620 "' E ~
(.!)
.E 1560 Cl 'iii :;::: OHL3
1500 DHL9
VHL15
•Al9 1440
1380
1320
2 3 4 5 6 7 8 9 10
Cycle Number
FIGURES Weights of lime-stabilized ashes exposed to wet-dry
cycles (28 days).
occurs in the stabilized mixtures, resulting in very high levels
of strength gain. Visual observations of the specimens also
indicated that none of the specimens experienced detectable
shrinkage cracking.
Results shown in Tables 3 and 4 further indicate, with few
exceptions, that the vacuum saturation test by using water produced
minor to moderate strength losses. Good durability may be achieved
with both lime- and cement-stabilized fly ash by adding a
sufficient amount of stabilizer and by pro-viding adequate curing.
Interesting results have been obtained with the acetic acid vacuum
saturation test. Exposure to acetic acid has caused higher strength
losses in lime-stabilized mix-tures than in cement-stabilized
mixtures, but strength gains are observed with the
cement-stabilized Harrison. This may be attributable to the
formation of cement-like ionic com-pounds, such as ferric acetate,
resulting from the dissolution of Fe20 3 present in the Harrison
ash . The same favorable effect as a result of acetic acid exposure
was also experienced with the cement-stabilized Harrison ash in the
later phase of the testing program involving permeability
evaluation. This was manifested as appreciable decreases in
permeability (see Bow de rs et al., this Record).
Vacuum Saturation Versus Freeze-Thaw
The vacuum saturation method is frequently used to evaluate the
freeze-thaw durability of the pozzolan-aggregate bases on
67
1960
1900
1840
1780
1720 "' E "' (3 .E 1660 Cl 'iii :;::: OHC3
1600 DHC9
VHC15
•AC9 1540
1480
1420
2 3 4 5 6 7 8 9 10
Cycle Number
FIGURE 6 Weights of cement-stabilized ashes exposed to wet-dry
cycles (28 days).
the basis of the excellent correlations obtained between the
results of the vacuum saturation (with water) and cyclic
freeze-thaw tests (14). The data obtained in this study were used
in correlation and linear regression analyses to assess whether
this would hold true for stabilized fly ash mixtures. Results
presented in Table 5 indicate that the correlations b tween the two
tests are fairly good, as reflected by the relatively high c
rrelation coefficients, and are excellent in the c;.i . e of
cement-stabilized a ·hes, where the c rrelation coefficient is 96.
7 percent. The regression equations in Table 5 represent the
statistical relationships between the residual strengths obtained
after vacuum saturation and 10 freeze-thaw cycles. On the basis of
these analyses, the vacuum saturation test can be used in lieu of
the freeze-thaw test to predict the freeze-thaw durability of
stabilized fly ash mixtures.
SUMMARY AND CONCLUSIONS
Limited applications and engineering properties related to the
stabilization of class F fly ash were discussed. The findings of
the studies lead to the following general observations and
conclusions.
1. Class F fly ash can be successfully stabilized with lime,
cement, or lime and cement combinations to produce a poz-zolanic
base course material that does not require the addition of
aggregate or soil.
-
68 TRANSPORTATION RESEARCH RECORD 1288
TABLE 5 CORRELATION AND LINEAR REGRESSION ANALYSIS OF
FREEZE-THAW AND VACUUM SATURATION RESIDUAL STRENGTHS
Curing Correlation Regression Mixture Period coefficient
Equation•
(days) r qFT = b+aqvs
Lime-stabilized ashes 7 0.815 qFT - 118.8 + 0.995 qvs
cement-stabilized ashes 7 0.967 qFT 72.3 + 0.858 qvs
Lime-stabilized ashes 28 0.863 qFT - 237.6 + 0.882 qvs
Cement-stabilized ashes 28 0.911 qFT a 141. 8 + 0.866 qvs
Stabilized Harrison ash 7 0.927 qFT - 172.7 + 0.733 qvs
Stabilized Amos ash 7 0.983 qFT - 73.3 + 0.977 qvs Stabilized
Harrison ash 28 0.844 qFT "' 133. 8 + l. ll qvs
Stabilized Amos ash 28 0.973 qFT D 176.7 + o. 716 qvs
*qFT residual strength after freeze-thaw test
qvs residual strength after vacuwn saturation test
a,b intercept and shape of the regression equation;
constants
2. The two fly ashes evaluated in this study exhibited high
levels of pozzolanic reactivity and satisfied all the relevant ASTM
specification criteria for pozzolans used in cement and concrete
and in lime-pozzolan stabilization. However, con-siderable
differences exist in the compaction , strength, and durability
characteristics of the stabilized fly ash mixtures .
3. In standard Proctor compaction, addition of increasing
percentages of lime to fly ash resulted in increased OMC and
decreased MDD for the mixtures . Addition of increasing
per-centages of cement did not affect the OMC and MDD
appre-ciably.
4. Studies indicated that adequate strength development and
durability levels can be achieved with stabilized fly ash by
incorporating sufficient amounts of lime or cement or both and
allowing the mixture to cure for a sufficient period. Achieving
adequate levels of strength before service exposure is
important.
5. In general, cement stabilization produced better strengths
than lime stabilization. For cement-stabilized fly ash. increas-ing
cement contents and extended curing resulted in increased strength.
For lime-stabilized fly ash, increasing lime content caused an
increase or decrease in strength, depending on the stabilizer
content and the length of curing. Extended curing , however,
increased strength invariably. The difference between the strengths
of lime- and cement-stabilized ash at a given stabilizer content
diminished somewhat as the curing period got longer.
6. Cement showed a dominant effect in strength develop-ment in
combined lime- and cement-stabilized fly ash mix-tures. Addition of
cement to partially replace lime markedly improved the early (7
days) and intermediate strengths (28 to 56 days).
7. Cement-stabilized fly ash mixtures, in general, exhibited
better durability characteristics than did the lime-stabilized
mixtures after being exposed to different environments. High
original (pre-exposure) strengths resulted in relatively high
residual strengths. With few exceptions, increased stabilizer
content and longer curing period enhanced durability.
8. Freeze-thaw cycles exposure produced the severest effects on
durability of both stabilized ashes and resulted in sub-stantial
strength losses . Wet-dry cycles. in contrast . did not have any
detrimental effect on durability and produced sig-nificant strength
gains without any shrinkage cracking. Inter-mediate effects were
observed rel ative to vacuum saturation with water or acetic acid
solution. Acetic acid had a positive effect (increased residual
strength) on durability with one of the cement-stabilized ashes.
indicating enhanced durability after exposure to an acidic
environment.
9. Results of the standard vacuum saturation tests (using water)
correlate very well with those of the cyclic freeze- thaw tests.
Therefore, vacuum saturation can accurately predict the freeze-thaw
durability of stabilized fly ash.
ACKNOWLEDGMENTS
The authors wish to thank Tareq Ashour and Peter Chou , graduate
research assistants at West Virginia University , who performed the
experimental work, and the Monongahela Power Company and American
Electric Power Corporation, who supplied the fly ashes.
Appreciation is also expressed to James S. Gidley for his valuable
contributions to the research project described in the paper. This
project was funded by the Energy and Water Research Center of West
Virginia University. This support is gratefully acknowledged.
REFERENCES
1. Lime-Fly sh Stabilized Bo cs ;mtl ubhase . Na1ional
oop-eraiivc Highway Research Program . y111Jiesi of Higf1 11111y
Pru -ticl! 14, HRB. Naii nal Resea rch Counci l , Washing1011.
1976.
2. E. J. Barenberg. Cement-Fly Ash Mixes for Pavements. Proc ..
Seventh International Ash Utilization Symposium. DOE/METC-85/6018.
Vol. 1. May 1985, pp. 57-66.
-
Usmen and Sowders
3. R. L. Terrel et al. Soil Stabilization in Pavement
Structures: A User's Manual, Vol. 2, Mixture Design Considerations.
FHWA-IP-80.2, FHWA, U.S. Department of Transportation, Oct.
1979.
4. J. F. Meyers, R. Pichumani, and B. S. Kapples . Fly Ash: A
Highway Construction Material. FHWA-IP-76-16, FHWA, U.S. Department
of Transportation, June 1976.
5. M.A. Usmen,J. J.Bowders,andJ. S.Giclley.LowPermeability
Liners Incorporating Fly Ash. In Disposal and Utilization of
Elec-tric Utility Wastes, (M. A. Usmen, ed.) ASCE , May 1988, pp.
50-65.
6. Fly Ash Design Manual for Road and Site Applications, Vol. I,
Dry and Conditioned Placement. GAi Consultants, Inc., EPRI, Feb.
1986.
7. L. J. Minnick. Reactions of Hydrated Lime with Pulverized
Fuel Ash. In Proceedings, Fly Ash Utilization Symposium, U.S.
Bureau of Mines Information Circular 8348, 1967, 1967, pp.
287-315.
8. H. B. Sutherland, T. W. Finlay, and I. A. Cram, Engineering
and Related Properties of Pulverized Fuel Ash. Journal of the
Institution of Highway Engineers, Vol. 15. June 1968, pp.
19-35.
9. D. H. Gray and Y. K. Lin. Engineering Properties of Compacted
Fly Ash . Journal of the Soil Mechanics and Foundations Division,
ASCE, Vol. 98, No . SM4, April 1972, pp. 361-380.
69
10. R. C. Joshi, D. M. Duncan, and H. McMaster. New and
Con-ventional Engineering Uses of Fly Ash. Transportation
Engi-neering Journal, ASCE, Vol. IOI, No. TE4, Nov. 1975, pp.
791-806.
11. D. C. Kraft et al. Ash Utilization in Bikeway Construction.
Proc., Fifth International A sh Utilization Symposium,
METC/SP-79/110, February 1979, pp. 694-712.
12. W. J. Head and R. K. Seals. Design of Experimental Haul Road
and Parking Lot Facilities Utilizing Power Plant Ash. Proc., Fifth
International Ash Utilization Symposium, METC/SP-79/JO, Feb. 1979,
pp. 679-693.
13. D. C. Oberst. Cement-Stabilized Fly Ash Road Base. Proc.,
Seventh International Ash Utilization Symposium. DOE/METC-85-6018,
Vol. 2, May 1985, pp. 590-596.
14. B. J. Dempsey and M. R. Thompson. Vacuum Saturation Method
for Predicting Freeze-Thaw Durability of Stabilized Materials. In
Highway Research Record 442, HRB, National Research Council,
Washington, D.C., 1973, pp. 44-57.
Publication of this paper sponsored by Cammi/lee on
Soil-Portland Cement Stabilization.